VTA glutamatergic projections to the nucleus accumbens suppress psychostimulant-seeking behavior

Abstract

Converging evidence indicates that both dopamine and glutamate neurotransmission within the nucleus accumbens (NAc) play a role in psychostimulant self-administration and relapse in rodent models. Increased NAc dopamine release from ventral tegmental area (VTA) inputs is critical to psychostimulant self-administration and NAc glutamate release from prelimbic prefrontal cortex (PFC) inputs synapsing on medium spiny neurons (MSNs) is critical to reinstatement of psychostimulant-seeking after extinction. The regulation of the activity of MSNs by VTA dopamine inputs has been extensively studied, and recent findings have demonstrated that VTA glutamate neurons target the NAc medial shell. Here, we determined whether the mesoaccumbal glutamatergic pathway plays a role in psychostimulant conditioned place preference and self-administration in mice. We used optogenetics to induce NAc release of glutamate from VTA inputs during the acquisition, expression, and reinstatement phases of cocaine- or methamphetamine-induced conditioned place preference (CPP), and during priming-induced reinstatement of cocaine-seeking behavior. We found that NAc medial shell release of glutamate resulting from the activation of VTA glutamatergic fibers did not affect the acquisition of cocaine-induced CPP, but it blocked the expression, stress- and priming-induced reinstatement of cocaine- and methamphetamine CPP, as well as it blocked the priming-induced reinstatement of cocaine-seeking behavior after extinction. These findings indicate that in contrast to the well-recognized mesoaccumbal dopamine system that is critical to psychostimulant reward and relapse, there is a parallel mesoaccumbal glutamatergic system that suppresses reward and psychostimulant-seeking behavior.

Fig. 1. NAc shell release of glutamate from VTA-VGluT2 fibers inhibits the expression and reinstatement of cocaine CPP.

A VTA injection of AAV5/2-DIO-eYFP or AAV5/2-DIO-ChR2-eYFP and NAc shell optic fibers. B VTA immunofluorescence detection of eYFP-expressing neurons (green) and TH (red). C eYFP fibers from VTA-VGluT2 neurons (green) in the NAc and optic fiber placements. D Cocaine CPP acquisition timeline (top). Both eYFP (n = 6) and ChR2-eYFP mice (n = 8) spent more time in the cocaine-paired chamber after the conditioning sessions (bottom; eYFP: chamber × experimental phase: F_2,10 = 6.05, p < 0.05; ChR2-eYFP: F_2,14 = 12.62, p < 0.001, ANOVA with Newman–Keuls post hoc-test). E Cocaine CPP expression timeline (top). ChR2-eYFP mice (n = 6) spent less time in the cocaine-paired chamber than eYFP control mice (n = 6) during photostimulation test (T). Both groups showed reinstatement of cocaine CPP during the cocaine-priming test (CT, bottom; eYFP: chamber × experimental phase: F_8,40 = 3.36, p < 0.01; ChR2_-eYFP: F_8,40 = 3.51, p < 0.01, ANOVA with Newman–Keuls post-hoc test). F Cocaine CPP reinstatement timeline (top). ChR2-eYFP mice (n = 9) spent less time in the cocaine-paired chamber than eYFP control mice (n = 8) during the cocaine-priming test (CT) paired with photostimulation (bottom; eYFP: chamber × experimental phase: F_8,56 = 6.92, p < 0.001; ChR2-eYFP: F_8,64 = 3.57, p < 0.01, ANOVA with Newman–Keuls post-hoc test₎. *p < 0.05, **p < 0.01, ***p < 0.001, against saline-paired chamber. Light-blue rectangles indicate photostimulation. ChR2 channelrhodopsin-2, Core nucleus accumbens core, DAPI 4’,6-diamidino-2-phenylindol, eYFP enhanced yellow fluorescent protein, mShell nucleus accumbens medial shell, TH tyrosine hydroxylase, VTA ventral tegmental area.

Fig. 2. NAc release of glutamate from VTA-VGluT2 fibers inhibits stress-induced reinstatement of cocaine CPP.

A VTA injection of AAV2-DIO-eYFP or AAV2-DIO-ChR2-eYFP and NAc shell optic fibers. B Stress-induced reinstatement timeline. C While eYFP control mice (n = 7) showed reinstatement of cocaine CPP during the stress-induced reinstatement test (RT) in the presence of NAc photostimulation, ChR2-eYFP mice (n = 7) avoided the cocaine-paired chamber during the stress-induced reinstatement test (RT) paired with NAc photostimulation (eYFP: chamber × experimental phase: F_6,36 = 3.03, p < 0.05; ChR2-eYFP: chamber × experimental phase: F_6,36 = 5.30, p < 0.001, ANOVA with Newman–Keuls post-hoc test). *p < 0.05, **p < 0.01, against saline-paired chamber. Light-blue rectangles indicate photostimulation. D Track plots from an eYFP (top) and a ChR2-eYFP (bottom) mouse during pretest (PT), expression test (T) and stress-induced reinstatement test (RT) in the presence of NAc photostimulation. Blue boxes indicate photostimulation.

Fig. 3. NAc release of glutamate from VTA-VGluT2-only neurons inhibits cocaine priming-induced reinstatement of cocaine CPP.

A VTA injection of INTRSECT viral vectors and NAc shell optic fibers. B Cocaine CPP expression and reinstatement timeline. C ChR2-eYFP mice (n = 8) spent less time in the cocaine-paired chamber than eYFP control mice (n = 8) during an expression test (T) paired with NAc photostimulation of VTA-VGluT2-only fibers. While both groups showed reinstatement of cocaine CPP during a cocaine-priming test (CT) in the absence of photostimulation, only ChR2-eYFP mice avoided the cocaine-paired chamber during a cocaine-priming test (CT) paired with photostimulation (eYFP: chamber x experimental phase: F_10,70 = 4.17, p < 0.001; ChR2-eYFP: F_10,70 = 5.24, p < 0.001, ANOVA with Newman–Keuls post hoc-test). D Track plots from an eYFP (top) and a ChR2-eYFP (bottom) mouse in which VTA-VGluT2-only neurons were targeted, during pretest (PT), expression test (T) and cocaine test (CT) in the presence or absence of NAc photostimulation. Blue boxes indicate photostimulation. E Both ChR2-eYFP (n = 8) and eYFP control mice (n = 6) preferred the cocaine-paired chamber during expression (T) and reinstatement (CT) tests paired with or without NAc photostimulation of fibers from dual VTA-VGluT2-TH neurons (eYFP: chamber x experimental phase: F_10,50 = 5.38, p < 0.001; ChR2-eYFP: F_10,70 = 9.34, p < 0.001, ANOVA with Newman–Keuls post-hoc test). F Both ChR2-eYFP (n = 6) and eYFP control mice (n = 8) preferred the cocaine-paired chamber during expression (T) and reinstatement (CT) tests paired with or without NAc photostimulation of VTA-TH-only fibers (eYFP: chamber × experimental phase: F_10,70 = 7.56, p < 0.001; ChR2-eYFP: F_10,50 = 7.06, p < 0.001, ANOVA with Newman–Keuls post-hoc test). *p < 0.05, **p < 0.01, ***p < 0.001, against saline-paired chamber. Light-blue rectangles indicate photostimulation.

Fig. 4. NAc release of glutamate from VTA-VGluT2 fibers inhibits instrumental reinstatement of cocaine-seeking behavior.

A VTA injection of AAV5/2-DIO-eYFP or AAV5/2-DIO-ChR2-eYFP and NAc shell optic fibers. B Cocaine self-administration and reinstatement timeline, showing training with cues and drug, extinction without cues or drug, and reinstatement conditions. Additional extinction sessions were run between each of the reinstatement conditions. C, D No differences were observed between ChR2-eYFP (n = 13) and eYFP control mice (n = 13) in the total number of lever presses (C, group × day × lever: F_9,216 = 0.23 p = 0.99, ANOVA with Newman–Keuls post-hoc test) or infusions obtained (D, group × day: F_9,216 = 0.29, p = 0.98, ANOVA with Newman–Keuls post-hoc test) during self-administration training. *p < 0.05, **p < 0.01, ***p < 0.001, against first day of training. E NAc photostimulation of VTA-VGluT2 fibers contingent to presses on the active lever during cocaine-priming test (CT) blocked cocaine-seeking behavior in ChR2-eYFP mice (n = 13) when compared to eYFP control mice (n = 13, group × day × lever: F_1,24 = 5.51, p < 0.05, ANOVA with Newman–Keuls post-hoc test). Cocaine priming-induced reinstatement of drug-seeking behavior in ChR2-eYFP mice was restored in the absence of VTA photostimulation (day × lever: F_1,24 = 35.55; p < 0.001, ANOVA with Newman–Keuls post-hoc test). **p < 0.01, ***p < 0.001, against extinction. +p < 0.05, against eYFP for the same experimental phase. Light-blue rectangle indicates photostimulation.

Fig. 5. NAc release of glutamate from VTA-VGluT2 fibers inhibits the expression and reinstatement of methamphetamine CPP.

A VTA injection of AAV5/2-DIO-eYFP or AAV5/2-DIO-ChR2-eYFP and NAc shell optic fibers. B Methamphetamine CPP expression and reinstatement timeline. C ChR2-eYFP mice (n = 6) spent less time in the methamphetamine-paired chamber than eYFP control mice (n = 6) during the expression test (T) paired with NAc photostimulation of VTA-VGluT2 fibers. While both groups showed reinstatement of methamphetamine CPP during the methamphetamine-priming test (MT) in the absence of photostimulation, only ChR2-eYFP mice avoided the methamphetamine-paired chamber during the methamphetamine-priming test (MT) paired with photostimulation (eYFP: chamber × experimental phase: F_10,50 = 4.27, p < 0.001; ChR2-eYFP: chamber × experimental phase: F_10,50 = 4.23, p < 0.001, ANOVA with Newman–Keuls post-hoc test). *p < 0.05, against saline-paired chamber. Light-blue rectangles indicate photostimulation. D Track plots from an eYFP (top) and a ChR2-eYFP (bottom) mouse during pretest (PT), expression test (T) and methamphetamine test (MT) in the presence or absence of NAc photostimulation. Blue boxes indicate photostimulation.